What's the fastest way to etch PCBs?

[cvriv]

I recently bought a 50watt JPT fiber laser. After learning all about it I finally came to the point that I was ready to laser engrave the mask for etching. I'm able to laser engrave both sides using alignment holes that I mill first. After the engraving I tried Ferric Chloride, but it takes way to damn long. I did sand the copper down before spray painting the clad. After 30 or so minutes a good amount of the copper was eaten away on one side but nothing on the other side. Some of the paint failed on the side with eaten away copper as well. What am I doing wrong here?

[Whales]

What temperature was the FCl?   Hotter is much faster.

Were you agitating the mix or leaving it sit?  For single sided boards you can get away with leaving it sit sometimes (but it doesn't always evenly etch).  For double-sided you definitely need to agitate.

It's been a while since I last etched a PCB, but I use hydrogen peroxide + an acid + salt instead of ferric chloride.

[cvriv]

I was etching at room temperature. I would agitate every few minutes or so. Im doing double sided. Im about ready to ready again after burning through most of the copper. Curious to see whats left after the burn. im assuming some of the paint is missing.

[cvriv]

Wow. OK. I heated up the ferric chloride and that worker much faster. I have to see what caused it to work much faster though. I chewed through a lot of copper with the laser.

[janoc]

Burning the copper with the laser likely leaves the surface oxidized/scorched and that will not etch well. Also, if the laser is actually melting the copper, then you are going to get metal deposits all over the place. Likely not what you want.

Seriously, if you want to make mask for PCBs, just laminate a sheet of dry resist on it and expose a transparency using a bank of UV leds (I have seen people try to use a low power laser for direct plotting on the resist but it is difficult to get right). Or do a toner transfer. Don't mess with a laser for this.

That 50W laser is much more useful for other things than this.

Finally, when etching boards yourself, make sure to not make the newbie mistake of not using copper pours - large areas of copper take ages to etch and use up a lot of your etching solution as well.

[Berni]

Ferric chloride is indeed very temperature sensitive.

At below room temp it basically takes forever. At room temp is can take something like 1 hour. But if you heat it up to something like 50°C is can do it in under 15 minutes.

The stuff does make a awful yellow mess tho. I find it the worst chemical for PCB etching.  Sodium persulfate is a much nicer chemical. It is a white non hydroscopic powder that doesn't make a mess and is faster too(but does also want to work heated warm). This stuff works best in bubbling tanks because it wants oxygen (But you can add a little bit of peroxide to give it a oxygen kick instead). It is also easier to see the board progress since the liquid is clear.

But if you want to go really fast the best way is to go the good-ol hydrocloric acid route. You can buy a liter of HCl and peroxide really cheep in the store. Just mix them with water in the proportions you want to tune how fast you want to etch. A warm strong mix will eat a PCB in a minute flat, while bubbling angerly and filling the room with acid vapors. You do not actually want to go that fast, it makes the process hard to control and run evenly, while making the devilish liquid even more dangerous than necessary. But in any case this is not something you want to do in the house, do it in a shed with good ventilation or outside.

As for oxides messing up the etch. Acids actually love to eat the oxidized state of metals, so with a lot of etching methods it should not be a problem. We do have a 40W fiber laser at work and i tried doing PCBs on it before. You can tune the laser settings just right so that it only burns the paint off but leaves the underlying metal shiny clean and untouched. I just used pre UV film coated PCBs, burned the UV film off and dunked it straight into enchant to get a decent board with 0.5mm pitch QFNs on it. I also tried to directly cut the copper traces into the board with the laser, but i never got that to work. It does have the power to burn away the copper layer, but once the laser is cranked up that high it also burns the FR4 under it. It is possible to do (including drilling holes), but i was not stubborn enough to figure out all the tricks.

[jmelson]

I have a Kepro spray etcher.  It has two titanium pumps that spray jets of FeCl out of a row of holes.  they suspend fiberglass windowscreen material to diffuse the spray jets, and it has a titanium heater that heats the etchant to about 45C.  When the board is properly cleaned and the resist is fully developed, it etches 1 Oz Cu through in less than 2 minutes.  I use Riston dry film photoresist from Think and Tinker.

Jon

[cvriv]

So I tuned my laser to just burn off the paint while leaving the copper untouched. At room temp the FC took way to long. I did the same thing but warmed the FC and it was faster but the paint mask came off.

I tried burning through most of the copper, and succeeded, but the heat destroyed some of the painted mask. I'm making TQFP 48 breakout boards for myself so the traces are very tiny.

I bought some high heat grill paint that can withstand up to 1200deg. Im wondering if this would help.

I'll look into the other chemical routes.

[kripton2035]

this should be possible with your 50W JPT laser ?

[cvriv]

Yea I can do that and have done it, but it's hard once the traces get very small and very close together. I working on a TQFP48 breakout board now. The traces are 0.3mm wide with 0.2mm spacing in some areas. The amount of heat need to blast through the copper will delaminate the traces in some areas. I'm also trying to remove all unwanted copper as well. I can merely just isolate the traces and that works fine, but I kind of want to see what it looks like with just the wanted traces. It's kind of a goal of mine. LOL. Im stubborn.

This was one of my very first attempts at lasering a PCB. I can't remember what the settings were that I used but it was way to much LOL. VOLUME WARNING.

[cvriv]

The results are in! This is a test run. Going to try the real board now. There's definitely some micro pitting going on here. I used high heat grill paint. It was actually a lot hard to laser off than regular paint so that stuff really does work. Maybe I'll try and another type of paint.

[free_electron]

this should be possible with your 50W JPT laser ?

what are we looking at here  ? 10 minutes to make the WRONG image. should have inverted it. now you have a no copper where the pattern should be , and copper where none should be ...
second question ? how much is that laser going to burn the fiberglass material ? and how conductive will it be after it has been scorched ....

10 minute ? a spray etcher can do that in a minute.

[cvriv]

If your laser settings are dialed in nicely you can easily burn through the copper isolating the traces you need without significantly burning the fiberglass. The board will be ugly but it'll work.

My third test didnt go well. The acid didnt eat though the copper nearly as fast. I made a new solution and still didnt eat through nearly as fast. I actually gave up. I sanded the clad off before painting it. I wonder if there was still some of the protective coating on the clad. There is a protective coating on the copper clad right?

[jpanhalt]

There is a protective coating on the copper clad right?

Not on ordinary bare FR4 that I  bought.  Take some of your fresh board, and put it in the ferric chloride.  It should quickly turn a salmon pink upon rinsing.  An alternative to FeCl3 is cupric chloride (CuCl2).  It's green and is easily made.  In my experience, it is a little faster but gives slightly less resolution.  That is, undercutting is worse and the end point is less forgiving.  I used photomasking for my resist. 

As for the paint you are using, have you considered using something with a dye rather than a pigment in in it?   

[janoc]

There is a protective coating on the copper clad right?

No, there isn't - unless you are using one covered with a photo-sensitive resist already, for whatever reason. Bare copperclads used for PCB manufacturing don't have any covering. They may be oxidized but the etchant should dissolve that rapidly. There is something wrong with your process.

I also don't see the point of "etching fast" - once the board etches within 10-15 minutes, what is the point of trying to "optimize" it to be even faster - at the expense of chemicals, danger (especially if using the concentrated acid/peroxide horror) and complex setup? You aren't really going into mass-production where every second counts with this, are you?

Especially when you aren't even using the right tools for fast etching, such as that spray etcher machine. Yeah, a bit spendy piece of kit for a home gamer but hey, you are the one messing with a fiber laser here ... Agitating the solution, e.g. by bubbling air through it (aquarium pump works) or by spraying a warm (not hot!) solution on the board continuously make a big difference because it ensures the reaction products get removed and a fresh etchant is constantly being supplied to the spots that need to be etched. If you just leave it sit or stir it by hand ever once in a while it will be slow.

[kripton2035]

this should be possible with your 50W JPT laser ?

10 minute ? a spray etcher can do that in a minute.

you didn't count the painting, uv exposure, the chemicals, the cleaning.
doing a pcb in one laser operation (even if it is 10 loops)  is really interesting (of course with the picture in the good way !)

[kripton2035]

Yea I can do that and have done it, but it's hard once the traces get very small and very close together. I working on a TQFP48 breakout board now. The traces are 0.3mm wide with 0.2mm spacing in some areas. The amount of heat need to blast through the copper will delaminate the traces in some areas. I'm also trying to remove all unwanted copper as well. I can merely just isolate the traces and that works fine, but I kind of want to see what it looks like with just the wanted traces. It's kind of a goal of mine. LOL. Im stubborn.
This was one of my very first attempts at lasering a PCB. I can't remember what the settings were that I used but it was way to much LOL. VOLUME WARNING.

yes... way too much !
it seems you can make very fine traces ... with the good setup !

[janoc]

Ferric chloride is indeed very temperature sensitive.

At below room temp it basically takes forever. At room temp is can take something like 1 hour. But if you heat it up to something like 50°C is can do it in under 15 minutes.

Depends a lot on concentration and whether the solution is fresh or not. But yes, warming it up does help (as with most chemical reactions). 1 hour is a bit extreme, with a fresh solution normal etching times at room temp are on the order of 20-30minutes.

The stuff does make a awful yellow mess tho. I find it the worst chemical for PCB etching.  Sodium persulfate is a much nicer chemical. It is a white non hydroscopic powder that doesn't make a mess and is faster too(but does also want to work heated warm). This stuff works best in bubbling tanks because it wants oxygen (But you can add a little bit of peroxide to give it a oxygen kick instead). It is also easier to see the board progress since the liquid is clear.

Persulfate is good but not great for home gamer that needs one board every once in a while. It absolutely needs warming because it does not etch at all at room temperature, the solution can't be reused and goes bad (decomposes, releasing oxygen - don't store it in an airtight bottle or it will burst!) quickly, so has to be prepared fresh each time. It does get pricey rather quickly like that.

From this point of view I have found ferric chloride a good compromise - it is inexpensive, lasts forever, can be regenerated ... Yes it does make a godawful mess if you spill it. That's the downside. So it is best to work outside (definitely not in a stainless steel kitchen sink) and take precautions to contain any accidents.

But if you want to go really fast the best way is to go the good-ol hydrocloric acid route. You can buy a liter of HCl and peroxide really cheep in the store. Just mix them with water in the proportions you want to tune how fast you want to etch. A warm strong mix will eat a PCB in a minute flat, while bubbling angerly and filling the room with acid vapors. You do not actually want to go that fast, it makes the process hard to control and run evenly, while making the devilish liquid even more dangerous than necessary. But in any case this is not something you want to do in the house, do it in a shed with good ventilation or outside.

Uuugh, no, not this devil's concoction, especially if using higher concentrations of acid and peroxide. That pretty much must be used out in the open - and far away from any metals because the fumes are not only deadly (literally) but will corrode anything metallic in range. And you certainly don't want to heat this - that reaction is exothermic by itself already (the acid acts as a catalyst decomposing the peroxide) and it could trigger a runaway reaction. That's an accident waiting to happen, with hot acid and peroxide furiously bubbling, boiling and splashing everywhere (+ the horrific fumes).

The only thing this has going for it is speed (but are we really in such a hurry to etch a board under a minute??) and it is very cheap if one can get a strong enough hydrogen peroxide (the common pharmacy stuff is too diluted for this to work well - and pricey!). That is often highly non-trivial these days because the thing is commonly used for things like drug making and bombs ...

[RoGeorge]

- heat warm the FeCl₃ for a faster etch (10-20 minutes)
- if it's single side copper clad, face the copper side down
- if it's too concentrated, it might work slower
- may use the sponge+zipbag method for very fast etch, low mess, very economic on chemicals
- may use a dedicated bubbles tank (with vertical sitting PCB) slower than the sponge method but faster than just letting the board in
- if the etching is in a tray, continuously agitating/moving the whole tray will speed etching a little

When we were kids we were using concentrated Nitric Acid (HNO₃), it's very fast (1-3 minutes to etch), but also very nasty to clothes, skin and it burns and bubbles everything it touches.  Generates very toxic fumes, can only be used outdoors.

[janoc]

When we were kids we were using concentrated Nitric Acid (HNO₃), it's very fast (1-3 minutes to etch), but also very nasty to clothes, skin and it burns and bubbles everything it touches.  Generates very toxic fumes, can only be used outdoors.

Yeah, that's the only common acid that will actually dissolve copper without help. But good luck getting it these days as a "mere mortal" without a chemistry lab access/procurement given how useful it is for making explosives ...

[cvriv]

I dont know where everyone got the idea that i wanted to etch in under a minute. 10 - 15 minutes is OK. The problem was that the etch was going over 30 minutes and not even fully etching.

Ive being trying the acid and peroxide, outside with a fan blowing the fumes out and away so im downwind from it.

Im getting mixed results here. The first two test boards etched nicely. In about 10 minutes. The next two test board did horrible. I used new fresh acid and peroxide and still wouldnt etch. I literally left them to sit over night and the traces of one are just gone... The the copper of the other is still there! They were both soaking in same solution. Just weird.

The only difference is the the first two boards were masked with a sharpie and the other with grill paint. The next two were masked with a regular paint + primer mix spray paint. In wondering if the paint did something to cause the acid to not work as well.

[cvriv]

And yes you can cut some super fine traces with the laser if you dial it in properly. The important thing is to try and keep the board as cool as possible so the copper doesnt delaminate.

[free_electron]

you didn't count the painting, uv exposure, the chemicals, the cleaning.
doing a pcb in one laser operation (even if it is 10 loops)  is really interesting (of course with the picture in the good way !)

buy presensitized boards. developing is 10 to 20 seconds.

and what with the fumes from that laser vaporizing copper ? that's gotta stink up the house like anything. i'll stick to my ammonium persulfate.

[cvriv]

you didn't count the painting, uv exposure, the chemicals, the cleaning.
doing a pcb in one laser operation (even if it is 10 loops)  is really interesting (of course with the picture in the good way !)

buy presensitized boards. developing is 10 to 20 seconds.

and what with the fumes from that laser vaporizing copper ? that's gotta stink up the house like anything. i'll stick to my ammonium persulfate.

I actually have two ventilation systems I use with my laser. I have a 4in inline fan with a hose to exhaust most of the cutting and an 8in fan that I use, as of right now, to vent the room if it starts to smell in the room.

[cvriv]

Getting closer. I used grill paint this time and the copper was eaten away within minutes. I was pushing the PCB up and out of the solution to let it breathe and I think that worked. I also lightly brushed the surface with a toothbrush as well. Some of the tiniest parts of the numbers got eaten away. Going to try again.

[james_s]

I use a mixture of hydrochloric acid and hydrogen peroxide when I want really fast etches, it takes only a few minutes, I typically warm it up first. Temperature makes a HUGE difference, all chemical reactions occur faster at elevated temperature. It should never take more than about 10-15 minutes to etch with any process if you're doing it right.

[cvriv]

Ive been etching outside were it's cooler. I dont want to bring it inside and don't have a means of heating it out outside.

[james_s]

I use a cheap aquarium heater, I think it was under $10 shipped from China. I usually etch out on the back deck with the acid and peroxide. Ferric Chloride doesn't smell so I used to use it indoors. You simply have to warm etchant up for it to work properly. In a pinch you can place your etching tray inside a larger tray of hot water.

[free_electron]

it could be nice to use a laser to do the exposure of the photoresist. no more need for films. essentially home LDI ...

[Berni]

Getting closer. I used grill paint this time and the copper was eaten away within minutes. I was pushing the PCB up and out of the solution to let it breathe and I think that worked. I also lightly brushed the surface with a toothbrush as well. Some of the tiniest parts of the numbers got eaten away. Going to try again.

Nice result there.

Yes agitation is very important in getting a good etch. This is why the real pro machines spray enchant at the board at high pressure.

Lifting it periodically helps because it makes the enchant run off the board, carrying off the used up liquid. scrubbing works even better since it mechanically carries away the liquid from the surface. It is only the tiny layer of liquid touching the board that is doing the work and due to viscosity effects it loves to stick to it and not move away( FeCl is especially viscous).

I personally gave up on etching once the Chinese started making these cheep boards for a few bucks. The boards they make are way better than anything i could make (like plated vias) with none of the mess. Tho it is annoying having to wait. I would make boards again if i could get the fiber laser to directly engrave it, that way i wouldn't have any of the chemical mess.

[kripton2035]

I personally gave up on etching once the Chinese started making these cheep boards for a few bucks. The boards they make are way better than anything i could make (like plated vias) with none of the mess. Tho it is annoying having to wait. I would make boards again if i could get the fiber laser to directly engrave it, that way i wouldn't have any of the chemical mess.

+++++1

[cvriv]

Most of the boards im going to make will probably be cut with the fiber laser because most of the boards ill need wont need to be this intricate. Ill be ok with leaving most of the copper on the PCB only having to remove enough for isolation. That is actually really easy to do. Also i can make my own stencils now to solder resist and paste as well. 

[Berni]

I am interested in how well cutting thin steel goes.

Gave it a quick try on the 40W fiber laser, it did cut it but also warped the heck out of it from thermal stress. Didn't fine tune the settings too much tho.

[chickenHeadKnob]

Yes agitation is very important in getting a good etch. This is why the real pro machines spray enchant at the board at high pressure.

Lifting it periodically helps because it makes the enchant run off the board, carrying off the used up liquid. scrubbing works even better since it mechanically carries away the liquid from the surface. It is only the tiny layer of liquid touching the board that is doing the work and due to viscosity effects it loves to stick to it and not move away( FeCl is especially viscous).

 Yes this ^. Warm the FeCl and get one of those cheap synthetic foam paint  brushes. No hard scrubbing required, just dab or brush the surface layer away.

[cvriv]

I am interested in how well cutting thin steel goes.

Gave it a quick try on the 40W fiber laser, it did cut it but also warped the heck out of it from thermal stress. Didn't fine tune the settings too much tho.

Yea it will probaby warp, but i havent tried it yet.

[mikeselectricstuff]

Fresh, warm (40-50 deg.C) ferric chloride with bubble agitation will etch 1 oz copper in about 10 minutes

[RoGeorge]

Sponge rub, 1 minute
https://www.instructables.com/Sponge-Ferric-Chloride-Method-Etch-Circuit-Bo/

Ziplock bag, less messy
https://rickettslab.org/bits2waves/fabrication-etching-pcb/

and a video, similar ideas:

[free_electron]

I personally gave up on etching once the Chinese started making these cheep boards for a few bucks. The boards they make are way better than anything i could make (like plated vias) with none of the mess. Tho it is annoying having to wait. I would make boards again if i could get the fiber laser to directly engrave it, that way i wouldn't have any of the chemical mess.

+++++1

<- this . no mucking about with chemicals or expensive lasers ( 8K for a basic fiber laser , and then you can do single sided. no soldermask , silkscreen , no metallisations ... drilling of holes ? ) and you dont stink up the house with noxious copper and fiberglass fumes.

delivered home, no fuss, silkscreen soldermask , plated thru holes. enig or hasl , no burnt-to-a-crisp-and-oxidised-so-bad-you-cant-solder-it pads.

[free_electron]

Sponge rub, 1 minute
https://www.instructables.com/Sponge-Ferric-Chloride-Method-Etch-Circuit-Bo/

Ziplock bag, less messy
https://rickettslab.org/bits2waves/fabrication-etching-pcb/

and a video, similar ideas:

spongebath fecl-pants ?

[cvriv]

I actually like the acid + perxoide solution. What i need to get are some cheap small paint brushes to brush the surfaces while etching.

[free_electron]

Fresh, warm (40-50 deg.C) ferric chloride with bubble agitation will etch 1 oz copper in about 10 minutes

time to switch to ammonium persulfate.
- no mess
- clean (transparent etchant that turns blue as copper dissolves. when it got the color of a pair of jeans it's time to make a new batch)
- can be regenerated
- cheap (go to a chemical products dealer. don't bother with the hobby shop.  years ago i bought a barrel (50KG) for about 200$. i split it with the local hobbyclub at 1/5th of what the electronics shop was asking. Sold half the barrel, so i'm sitting on 25Kg's of free etchant. i think i'm about halfway through the 25Kg. that was 30 years ago, somewhere late 80's early 90's   i'm getting old. i'm still thinking 1990 is 10 years ago ... )

con:
- needs to be run warm (40 to 50 degree C )
- it does eat holes in natural fiber clothes. any splash and it's too late. even if you wash it out immediately the holes will develop over time. then again having fecl stains is a sure way to have to scrap your clothes too.

if you have a spray etcher, this stuff will eat a 18x24 inch doublesided panel in under a minute. It's what they use in the real PCB fabs. FECL is long gone.

[janoc]

if you have a spray etcher, this stuff will eat a 18x24 inch doublesided panel in under a minute. It's what they use in the real PCB fabs. FECL is long gone.

That is not quite true. I don't remember which one but either JLCPCB or PCBWay was clearly showing FeCL3 etching in one of their factory tours. It is still a rather common process because it is both cheap and the chemicals can be reused. Moreover, one doesn't care about mess or that the etchant isn't clear on an automated line where the mess is contained inside of the machine anyway.

Persulfate is nice - if you can get it cheaply and thus can afford to make a new batch every time because it doesn't store well (it decomposes). Also not quite sure about the ability to regenerate it, given that it decomposes during etching ...

[janoc]

it could be nice to use a laser to do the exposure of the photoresist. no more need for films. essentially home LDI ...

Some people have tried this but it is very difficult to both focus the laser accurately and to adjust the power down to be just perfect for the resist (laser is very likely overexpose the resist due to how bright it is). The project I have seen had also problems with vibrations preventing really fine details from being exposed but that was because they used diode laser and a plotter/gantry like setup, not galvos (which would be a different price/complexity league).



Probably a more feasible way to do this would reusing the laser printer with the scanning mirror mechanism. This has also been done:


But apart from the fun/tinkering aspect I don't see the point - DIYing a board that requires this sort of setup and time is probably not worth it. Simple stuff can be done with toner transfer easily and complex, very fine setup -> outsource it to a fab.

[free_electron]

if you have a spray etcher, this stuff will eat a 18x24 inch doublesided panel in under a minute. It's what they use in the real PCB fabs. FECL is long gone.

That is not quite true. I don't remember which one but either JLCPCB or PCBWay was clearly showing FeCL3 etching in one of their factory tours. It is still a rather common process because it is both cheap and the chemicals can be reused. Moreover, one doesn't care about mess or that the etchant isn't clear on an automated line where the mess is contained inside of the machine anyway.

Persulfate is nice - if you can get it cheaply and thus can afford to make a new batch every time because it doesn't store well (it decomposes). Also not quite sure about the ability to regenerate it, given that it decomposes during etching ...

fecl is not usable on tin mask. Any board with plated holes uses tin mask. Maybe on some cheapo single sided board you can still use fecl. i don't know of any of my fabs that still use it.

Regeneration is done by cooling the etchant to around 0 degree C . This crystallize the metal sulfate without touching the remaining ammonium persulfate. The crystals are then removed (centrifuge) and the remaining etchant is sent back into the etcher and topped off with fresh ammonium persulfate.
The metal sulfate crystals are then sent for extraction of the copper.

There are two base principles for etching circuit boards :
- acidic (Ferro chloride and copper chloride) or alkaline (Persulfate)

pro's and cons:
Acidic :
PRO Acidic is precise but slower than alkaline
PRO acidic does not react with the photoresist
PRO acidic minimizes undercuts so you get a more vertical wall (this is due to the etch rate)
PRO CuCL is used as etchant (NOT fecl !)
PRO CUCL has a constant etch rate, FECL doesn't
CON : Fecl is not compatible with tin resist so not usable for outer layer etch on double or multilayer !
CON : FECL is expensive to get rid off !
CUCL needs addition of HCL and extraction of chlorine gasses . chlorine gas destroys machinery and humans... so you need extra safety measures.
CON : regeneration is electrolytic and requires special macinery , high currents

Alkaline:
Pro : cheap
Pro : compatible with tin mask ( multilayers and double sided. anything that uses plated holes ) . you could use CUCL for the core , but why mess with two chemistries in the fab ?
PRo : easy regeneration (cool and mechanical separation -> send out for recycling)

[Berni]

That is not quite true. I don't remember which one but either JLCPCB or PCBWay was clearly showing FeCL3 etching in one of their factory tours. It is still a rather common process because it is both cheap and the chemicals can be reused. Moreover, one doesn't care about mess or that the etchant isn't clear on an automated line where the mess is contained inside of the machine anyway.

Persulfate is nice - if you can get it cheaply and thus can afford to make a new batch every time because it doesn't store well (it decomposes). Also not quite sure about the ability to regenerate it, given that it decomposes during etching ...

Yeah PCB factories still use FeCl a lot because of how easily it can be regenerated and reused. When hot and blasted at pressure it still is plenty fast.

The point with persuflate is that the eating away of metal consumes oxygen, but you can bubble air trough it to bring it more oxygen. I got a whole kilogram of sodium persuflate pretty cheep, so i never bothered reusing it. I just mixed up the minimum amount to fill a small flat plastic pan. Plenty still left that i will likely never use.  I think it can also be used as a oxidizer in rockets, so might try that some day when i m bored enough. I much preferred using it over FeCl (Especially since i bought FeCL in granule form) since it's less messy and worked better when just thrown into a pan, no fumes either.

Thing is that the dissolved copper afterwards is bad for the environment.

[fourfathom]

You can also make simple non-plated-through boards using a CNC mill.  LPFK makes a system for this, and I've done it on my Shapeoko mill, but the precision I can achieve is pretty poor.  Realistically, my setup is only useful for simple single-sided boards, perhaps with a groundplane.  Otherwise the setup is too tedious, and I'm lucky if I can run one trace between 0.1" DIP pads.

For anything complicated I use JLCPCB.  If I can wait a couple of weeks the shipping cost is quite low -- I recently got two small (1" x 2") two-layer board designs done, 5 boards each, delivered to the USA for a grand total of $12.80 -- which boggles my mind.  Four-layer boards are barely more expensive.  Fast shipping is more expensive, but still much less than my time is worth to me.

But I do understand the fun of Doing It Yourself -- There are lots of things I do that make no sense, except that I want to see if I can figure it out.

[free_electron]

Yeah PCB factories still use FeCl a lot because of how easily it can be regenerated and reused. When hot and blasted at pressure it still is plenty fast.

not true. see my post above. the process is not compatible with tin mask so you cannot make anything that uses plated thru holes. double sided and multilayer are impossible with Fecl.
Fecl is expensive to dispose of due to regulations.

Typically CuCl is used, not Fecl. but then only for inner layers. the outer layer needs alkaline process.
alkaline has a much higher etch rate than cucl or fecl.
large fabs may use both cucl and persulfate

[purfield]

I've been able to make laser-etch PCBs that look better than this.  I used a vanadium Q-switched laser with ~20ns pulses, but that probably doesn't make a huge difference.  I used the LPKF method of segmenting the large areas of copper removal to thermally isolate them and then defocus or wobble the laser to heat the segments until they delaminate.  The FR4 substrate looks much better that way. 

[rob77]

Fecl is expensive to dispose of due to regulations.

FeCl3 (there is no such thing as FeCl) is used also in sewage water treatment , so FeCl3 alone shouldn't be such a big deal. i would assume CuCl2 in the used-up etchant is a bigger problem.

[rob77]

Yeah PCB factories still use FeCl a lot because of how easily it can be regenerated and reused. When hot and blasted at pressure it still is plenty fast.

alkaline is even faster and it doesn't eat tin (important for double sided through hole process) that's why PCB houses are using alkaline... in fact you produce more etchant during the etching itself you just need to add some ammonia here and there - it doesn't need a separate regeneration process as used-up fecl3 does.

[cvriv]

Tinned and grill painted. I have UV resist, but im impatient plus i want to see how the grill paint holds up while soldering.

[josuah]

That looks rather good! How was the masking made before etching? With laser-engraved mask as you first described? The laser burning part of the mask that should not be there?

I am fed-up with toner transfer: depends on a combination of too many parameters in my experience.
I have even tried heating acetone-soaked paper with toner as I stomped on it as I kissed a rabbit's paw with 777 four-leaved clovers in the mouth...

Then I see people posting videos of single-handedly transferring toner without paying much attention and getting a perfect result...
How are they doing?

I am switch to UV sensible mask TONIGHT!